AI Armageddon is coming. Why aren’t we panicking? – The Mail & Guardian

Artificial intelligence will replace most jobs and create few. The future is bleak – except for the rich. (Science Photo Library/ Sergi Laremenko)

For the past few months, I’ve been experimenting with OpenAI’s ChatGPT o3 and 4.5 advanced reasoning models. 

The difference between these models and ChatGPT 3 is astounding. Grok 4, despite its infamous “MechaNazi” moments, is utterly incredible. With the right prompts, AI can do your taxes, respond to your emails, increase the efficiency of academic research and create deep fakes so realistic they’re almost indistinguishable from reality (and soon will be). They are imperfect. ChatGPT remains a terrible philosopher when left to its own devices, but this is just the beginning.

As AI models learn to train and improve themselves, the improvements will become exponential. AI super-intelligence is just around the corner — intelligence that far surpasses the capabilities of human beings. This is not science fiction. It’s real, and it’s now. AI models will go from what we have now to having capacities we can only imagine, over just a period of months.

AI will probably become a better, faster, cheaper architect than any human. It will render administrators, assistants, paralegals, data inputters, most programmers, most chemists, physicists and actuaries redundant. It will create artworks, novels and music in minutes. There will be AI teachers that can interact with students in a deeply personalised way. There’s a good chance they’ll be just as effective as real teachers. They’ll make no mistakes, and everyone gets one-on-one tuition at a fraction of the cost. Most research and development will be outsourced to the far more capable AIs.

This might have some benefits. Energy prices could tend to zero as AI discovers increasingly efficient means of energy production. AI in healthcare will enable us to predict pathologies years in advance, provide flawless diagnoses and personalised treatment plans, and help develop gene editing and other new forms of treatment. 

But the potential cost of all this is unimaginably awful, unless AI development is strictly regulated — and this seems very unlikely in the current political environment.

The first hint of the negative effects of AI will be an uptick in unemployment, as companies around the world (mainly in the West at first) start replacing the intellectual workforce with AI. I think we’re already seeing this (in its infancy). The revisions in the US job report this month might be a partial reflection of that. But over the next few years, there’s a good chance unemployment becomes a genuine global crisis. 

What happens then? Well, those left without jobs will be unable to pay their bonds on the properties they bought in the midst of a property bubble. They’ll go into negative equity, have their houses repossessed and be plunged into debt and bankruptcy. Demand for property will slump, and property prices will crash. 

In the West, a huge percentage of people’s wealth is tied up in property, so this inevitably leads to a global financial crisis. The response to arguments like this often runs as follows: this is no different from the industrial revolution. New jobs will be created, and it will lead to progress. But I think this is a disanalogy. The industrial revolution targeted physical labour. We still had intellectual labour to fall back on. The AI revolution takes away intellectual labour. Although it’s no doubt true that a few jobs will be created, where people specialise in applying AI technologies, these will be few and far between relative to the jobs lost. I just don’t see it. When the truck-driver or taxi-driver’s job is taken by an autonomous vehicle, is that person meant to become an AI entrepreneur? This response is detached from reality.

South Africa is used to unemployment. To mitigate the risk of unemployment, students come to universities to get degrees, improving their chances of employment. But once intellectual labour is replaced by AI, many of the degrees aren’t going to help. Many potential students will probably choose not to go. But without students, universities collapse. Without universities, the skill-set and critical thinking capacities of a population drop, and without critical thinking capacities, we become more vulnerable to political manipulation through AI-driven social media platforms and deep fakes.

In addition, there’s little stopping people from using AI to target the banking system, design killer viruses or infiltrate the Pentagon to start a nuclear war — truly the stuff of Armageddon movies.

Right now the global political leadership is devoid of a moral backbone. In exchange for political funding and (at least in Donald Trump’s case) removing “woke content”, Big Tech companies are given the freedom to fight the AI arms race almost unregulated. Super-intelligence will come so fast that society has no time to adjust. This is in the politicians’ interests, because they can use this technology to manipulate the electorate. 

Everyone with spare capital is throwing their money into NVIDIA and other AI-related stocks, delighted to see their portfolios skyrocket. But we’re not seeing the bigger picture. Sam Altman — chief executive of OpenAI — is. He’s established “Worldcoin”, a system linking crypto to biometric identity, with the main purpose of facilitating universal basic income. Essentially, he’s planning to further profit from the quasi-Marxist dystopia his company is driving us towards. If the guy driving OpenAI is envisioning a world where UBI becomes a global necessity, why aren’t we more worried?

Generally I’m a positive chap. I want to believe AI technology will be progressive. I’m sure universities will adapt and drive courses focused on critical thinking and other skills needed in this new world. Further, if AI can help us produce power basically for free, and use robots for all physical labour and AI for all intellectual labour, there’s a conceivable reality where nobody has to work, and our physical and psychological needs can be met with no money at all. 

But we live in a capitalist world led by people like Trump. OpenAI started as a nonprofit, open platform. That has changed. Profit matters more to Big Tech than population wellbeing. Sure, AI will lead to incredible advances in medical technologies — but who will see these benefits? My guess is the rich, and the rich alone. 

I hope I’m wrong, but it seems to me that a truly dystopian future might be right in front of us, and nobody is doing anything about it. There is no #juststopAI movement. Demands for global regulation exist, but they’re not mainstream. Where are the marches? Where are the protests? Why are we not putting pressure on our politicians to be more responsible? 

I urge you to consider what your plan will be in this new AI world. Choose your degrees wisely. Equip yourselves with critical thinking skills (I recommend philosophy!), learn how to use AI technologies (that should help, at least for a while), or consider industries where the labour force is less likely to be affected (such as health and basic education). Good luck, everyone.

Professor Benjamin Smart is a director at the Centre for Philosophy of Epidemiology, Medicine and Public Health at the University of Johannesburg.